Fingerprints, an example of which appears in FIG. 1, are a powerful biometric because of the uniqueness and stability of the feature signature of the human fingerprint pattern. The last decade has seen an improvement in the ability of automated image analysis systems to extract features from fingerprint patterns quickly, accurately, and consistently. All of these factors have contributed to the success of fingerprints as a biometric for person identification systems.
Person identification using fingerprints involves several steps. A first step involves fingerprint image acquisition, with the goal being the accurate reproduction of the fingerprint pattern in digital image form. A subsequent step is the accurate extraction of fingerprint features, known to be unique for every individual, from the digital image. In a later step, the pattern of these features is used to search through a database of patterns to determine the optimal match, and hence, to identify the correct individual.
The features used for identification are known as Galton details or minutiae, and relate to the location of points centered on specific patterns formed by the ridges, which appear as black lines in the fingerprint image of FIG. 1, and valleys, which appear as white lines. The comparison of the relative positions of these points with a reference fingerprint determines the degree to which a given unknown pattern matches the reference fingerprint.
There is a variety of different types of minutiae. FIG. 2 shows four of the most common types, isolated by encircling in white: A) Island; B) Dot; C) Bifurcation; and D) Ending Ridge. An average fingerprint may have 20 to 40 minutia points, although fingerprints of poor quality may have as few as 3 or 4, and fingerprints that are “rolled” to imprint more surface area of the finger may have as many as 100 or more. The specific two-dimensional layout of the minutia points uniquely characterizes an individual. Clearly, the more minutia points that are correctly located, the greater the probability that a given fingerprint will be accurately matched against its reference fingerprint in a given database. The goal in biometric systems is to maximize that probability; therefore, accurate feature extraction is central to this goal.
The examples of FIG. 2 seem to be quite clear. However, in practice, fingerprint images are significantly more degraded than those shown in FIG. 2 and the location (and even the type) of minutiae much more ambiguous. FIG. 3 shows examples of real minutiae and demonstrates the inherent uncertainty in an image, leading to problematic decisions for both experts and automated algorithms. For example, in FIG. 3, the magnified area 2 contains numerous points in the image that might be considered as ending ridges or dots. However, expert examination in the magnified area reveals only two minutiae: minutia point 4, which is a definite bifurcation 6; and minutia point 8 that, despite careful examination, is ambiguous, especially if the larger context of a ridge flow is considered. In fact, minutia point 8 can be considered a bifurcation 10 or an ending ridge (or even dot) 12 depending on whether the break in the arm is due to poor image quality or due to a real physical break. In other words, it is unknown whether the break is a physical characteristic of the fingerprint or an artifact of the image acquisition/analysis process because of inherent uncertainty in the image itself.
In instances of ambiguity, minutia extraction algorithms may do one of four things: 1) correctly locate a true minutia, yielding optimal results; 2) fail to locate a true minutia, which can weaken the probability of a subsequent match; 3) incorrectly locate a false minutia, which can later confuse the matching algorithm; or 4) correctly locate a true minutia but misidentify the minutia type, resulting in a mirror position offset that can sometimes appear as a missed true minutia and an incorrectly located false minutia (a hybrid of (2) and (3)—both weakening the probability of a match and confusing the matching algorithm).
The acquisition of fingerprint images of fingerprints can occur in many different ways. Irrespective of how the image is acquired, however, the image formation process is known to result in an inherently flawed recreation of the actual fingerprint. The flawed recreation occurs because of several reasons:
1. Image Deformation. Most fingerprint scanning devices require the subject either to press their finger onto a platen or to brush their finger against a scanning device. Because of the elasticity of skin and varying quantity of finger pressure, in both cases the fingerprint pattern can be slightly deformed, with slightly different deformations with each image acquisition.
2. Image Superposition. Dirty or oily fingers can leave behind residual fingerprints on a platen. If the platen is not cleaned between scans, as is too often the case, images of these residual prints can superimpose themselves on the scanned fingerprint.
3. Image Distortion. The image acquisition process relies on some method that measures the physical differences between ridges and valleys. Whether that means measuring capacitance, reflected sound, reflected radiation emitted radiation, or the like, the projection of an irregular 3-D object (the finger) onto a 2D flat plane inevitably introduces image distortions.
4. Image Resolution. Because of the relatively small size of minutiae, image resolution is a critical factor in facilitating accurate automatic detection. Higher resolution can yield superior definition, but in practice, resolution is often limited by cost and the technology available.